The long-term objective of this research is to understand both neural mechanisms for processing communication sounds and fundamental neural coding mechanisms in auditory cortex that subserve cortical representations of biologically relevant sounds. We will use the common marmoset (Callithrix jacchus) as our experimental model to address these questions. This model system provides several important advantages over other species, namely, a hearing range similar to that of humans, a rich vocal repertoire, an auditory cortex that lies largely on the lateral surface of the cerebral cortex and a high reproductive rate while in captivity. In this application, we will focus on elucidating information processing mechanisms in the rostral areas outside the primary auditory cortex (A1). Aim 1 will study neural representations of marmoset vocalizations in the rostral areas using virtual vocalization stimuli that we have recently developed in our laboratory. These stimuli are based on the statistics of marmoset vocalizations and can be easily manipulated in both spectral and temporal domains to probe cortical responses with great flexibility. Aim 2 will investigate the roles of spectral and temporal pitch mechanisms in generating pitch-selective neural responses in a pitch- region located in the rostral areas. Results of this aim will pave the way for further studies to investigate anatomical connectivity of pitch-selective neurons in auditory cortex. Aim 3 will use sleep as a unique behavior state to study the state-dependent processing in the rostral areas. We will quantitatively evaluate neural responses in the rostral areas to external sounds during sleep. The transformation of cortical representations of sound-evoked responses during sleep from A1 to the rostral areas will provide further insight into information processing streams within the superior temporal gyrus.